The world is certainly warming. What about where I live?

This post was written for Climatesnack where you can find this post as well as many others by other early-career climate scientists who like to write.

You may have heard that the world is getting warmer. But it would be perfectly fair to ask – what does that mean for my neighborhood?

I can be fairly confident, even without knowing exactly where you live, in telling you that it is also going to get warmer in your neighborhood, at least on average. But how much warmer? That is a harder question. Climate models do of course predict the temperature at every point in space, but the problem is, the smaller the region you look at, the more uncertain the temperature predictions become.

A climate modeling study by Clara Deser and coauthors [1] provides an illustrative example. They run the same climate model 40 times, with tiny differences in the initial state of the atmosphere. These tiny differences immediately set each of the 40 model runs down a very different path, an effect called the butterfly effect due to the ability of the mere flap of a butterfly’s wings to change the course of climate history.

Averaging the predicted temperature rise (from 2005 to 2060) over all models, gives a picture of how temperatures will change, shown below for North American winter. No one simulation looks just like this, and neither will the real world. The simulations with the warmest and coldest North Americas are also shown. They are very different from the average. Focusing on a few individual cities makes this even clearer. Next to the maps, you can find the trajectory of winter temperature for Seattle, Phoenix, Mazatlán, the entire United States, and the entire globe. Each plot shows the observations so far (black) and the largest (red) and smallest (blue) predicted winter temperature change out of the 40 model simulations.

Change in winter (December-February) temperature from 2005 to 2060 over North America in realizations of the CCSM3 climate model. Shown are the average change and the simulations with the warmest and coldest North American temperatures. The side panel shows for 3 cities, for the US mean, and for the global mean, the simulations with the largest (red) and smallest (blue) predicted temperature changes. Observations from 1910-2008 are shown in black. From Deser et al. 2012. [1]

Looking at a single city, there is something causing large fluctuations in time and large uncertainties between model simulations. Seattle winters could be 4°C warmer by 2060, or they could be slightly cooler. Not exactly a forecast you can rely on. There is even a similarly large uncertainty for the entire United States. Looking at the global average however, much of the uncertainty is gone, and this model gives an estimated global temperature rise, in winter, of 2-2.5°C. Something is distributing the same total amount of global warming to different regions at different times and in different realizations of the same model. The rest of this piece will explore why.

For those living in the midlatidudes, that is north of about 30° N (Houston, Cairo, and Central China) or south of about 30° S (Porto Alegre, Brazil, and a fair bit south of Johannesburg and Brisbane), nearly all fluctuations in weather result from large meanders of the jet stream called Rossby waves. These meanders are as big as continents. At any one time there are typically about 5 or 6 wave crests around the entire globe, an alternating pattern of northward and southward wind.

The Jet Stream

The jet stream is a band of persistent eastward winds in the midlatitudes. These winds blow at 4-10 m/s at the surface and 25-50 m/s in the upper atmosphere. They cause airplane flights to take longer from East to West and cause typical weather patterns to blow in from the West.

Due to the peculiar physics of rotating planets, these waves always move westward, the opposite direction of the jet stream. Most waves cannot move upstream fast enough to avoid being swept along to the east by the jet stream, hence why our weather usually comes from the West. But the biggest, fastest waves are able to move upstream at a speed equal to that with which the jet stream is pushing them, so they stick in one place, leading to northward or southward winds that persist in a region for weeks, months, or years.

These stationary Rossby waves tend to take similar patterns year after year because they interact with regions of warm ocean water, which are fairly consistent from year to year, and with mountain ranges, which, as far as the climate is concerned, do not move at all. Regions that get consistent poleward wind are relatively warm and wet (like Northern Europe and Alaska) and regions that get consistent equatorward wind are cold and dry (like northeastern Canada and Siberia). It turns out that it’s all just a matter of which way the winds are blowing.

East-West Variations in Temperature

If you don’t believe me that Northern Europe and Alaska are relatively warm, consider that cities such as Bergen, Norway and Anchorage, Alaska are at a similar latitude (60-62° North) to Yakutsk, in Siberia, where average winter temperatures are around -40°C, and to nearly uninhabited Nunavut Canada.

In Early 2014, the stationary wave pattern was a bit different from ‘normal’ with consequences all around the Northern Hemisphere. California experienced a record-breaking drought; the Eastern US received a harsh cold winter. In Central Europe, winter hardly seemed to happen at all. The winds came from the South, dumping above average snowfall in Italy and the Southern Alps and leaving Switzerland and Germany dry and warm. This was all part of a Rossby wave that extended around the Northern Hemisphere and persisted for most of the winter. This is exactly the kind of event that would cause a big spike toward colder temperatures in the United States, while leaving the global mean temperature relatively unchanged.

Returning to the 40 nearly identical climate simulations, we can see the role the Rossby waves play in regional climate uncertainty by adding the sea level pressure change to the temperature change plots. Sea level pressure variations are the surface manifestation of a Rossby wave extending throughout the atmosphere. The figure below, from another recent study by Clara Deser and collaborators [2], shows the temperature and precipitation change for two example climate simulations. Superimposed are the sea level pressure changes, where I have added the typical weather map arrows showing the wind direction associated with high and low pressure systems. The total change in temperature, precipitation, and sea level pressure is split up into two components: the average over the 40 simulations, which is the portion of the change forced by greenhouse gasses and the portion due to internal variability of the climate system associated with Rossby waves.

Map of the change in winter surface air temperature (shaded, °C), winter precipitation (shaded, mm/day), and sea level pressure (contours, 1 hPa, negative values are dashed) from 2005 to 2060. Two of the 40 realizations of the CCSM climate model are shown. Modified from Deser et al. 2014. [2]

Several differences between these simulations can be understood in terms of the difference in stationary Rossby wave pattern from average (shown in the middle). In C29, a high-pressure system is stuck off the coast of Alaska. The associated southward winds over western North America bring cold, dry conditions to Alaska, Western Canada, and the Pacific Northwest. The simulation C6 is nearly the opposite. A low-pressure system stuck off the coast of Alaska causes northward winds intersecting the west coast of North America. This brings warm wet conditions to the coast and warm dry conditions to the other side of the Canadian Rockies. Both simulations also have the change forced by greenhouse gasses, shown on the right. The main reason the temperature and precipitation change is not uniform from east to west is the small stationary Rossby wave change in response to greenhouse gasses.

The difference between these two simulations, due to the internal variability of the climate system, gives large uncertainties in the predicted change in temperature and rainfall over much of North America, especially on the west coast. The important thing to understand about this uncertainty in regional climate is that it would exist even if we had a perfect climate model, because it arises from tiny differences in initial conditions and we can’t measure the state of the atmosphere exactly. But this doesn’t mean we know nothing. Based on this model, we can say that, for example, there is a 68% chance that the San Francisco area receives on average, between 0.3 and 0.5 mm less rainfall per day, and is, on average 0.5 to 2.5 °C warmer, in 2060 than in 2005 (these are just rough approximations of the numbers for illustrative purposes). By averaging over larger areas or longer time periods, this uncertainty is reduced. Based on the recent IPCC report [3], there is a 90% chance that the global mean surface temperature, will be on average 0.8 - 1.8 °C warmer in 2050 than the 1986-2005 average.

The IPCC Report

The IPCC report is the detailed report of the Intergovernmental Panel on Climate Change on the physical basis, impacts, and mitigation of climate change. While it is not light reading, there are some good summaries for policymakers (full report and summaries on ipcc.ch). You can also get a shorter, more artistic summary from Gregory Johnson’s IPCC Haiku.

This presents a challenge in preparing for and convincing politicians of the dangers of climate change, because the regional changes that are ultimately what affect people have large uncertainties. But by communicating the range of possible future climates, climate scientists can help policy makers, city planners, and engineers prepare for all that climate change can throw at us. The IPCC report is already doing just that. What I hope you can take away from this post in particular, is that some of the uncertainty in future regional climate comes not from the failure of climate modelers, but from the irreducible uncertainty of a chaotic system, in particular with regards to the meanders of the jet stream we call Rossby waves.

References

¹Deser, C., R. Knutti, S. Solomon, and A. S. Phillips, 2012: Communication of the role of natural variability in future North American climate. Nat. Clim. Change, 2, 775-779

²Deser, C., A. S. Phillips, M. A. Alexander, and B. V. Smoliak, 2014: Projecting North American climate over the next 50 years: Uncertainty due to internal variability. Journal of Climate, 27(6), 2271-2296.

³IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp.